720
,. Figure 5. Twist-envelope conformation for ethylene phosphites;
Newman projection down C-C bond.
-
-
we can then assign HA(downfield, J p H a 2 cps) as cis to X and H B (upfield, J p H e 9 cps) as trans to X. However, the uncertainty about the conformational size of S and P lone pairs makes this a tentative conclusion. If 14 were favored rather than 15, the reverse assignment would be required. Structure of Ethylene Phosphites. The vicinal hydrogen-hydrogen coupling constants are not in agreement with a simple envelope conformation (15). Instead, 14 and/or 15 must be twisted to yield an
H-C-C-H dihedral angle of about 30". Therefore, taking 15 to be the favored envelope form, ethylene phosphites appear to prefer two equivalent twist-envelope forms, one of which is approximately shown in Figure 5 . A time average of other conformations yielding the same effect as H-C-C-H dihedral angles of about 30" and very different P-0-C-H dihedral angles would, of course, also satisfy the results found here. However, it seems quite reasonable that the nonbonded interactions in ethylene phosphites and ethylene sulfite should result in the twist-envelope conformation being favored. This conformational result was obtainable because both H-C-C-H and P-0-C-H couplings - could be observed. However, the twist-envelope conformation may well have applicability to other five-membered rings. Acknowledgment. The 60-Mc spectra were taken on a Varian A-60 instrument purchased by an institutional grant from the National Science Foundation. The 100-Mc spectra were taken on a Varian HA-100 instrument at the University of California at Riverside through the courtesy of Professor Robert C. Neuman. We thank Professor H. Goldwhite for a Pal-decoupled spectrum of 10. The machine calculations were done on an IRM 7094 at the UCLA Computing Facility. Valuable conversations with Drs. F. A. L. Anet, J. L. Sudmeier, and P. C. Turley are acknowledged.
A Kinetic Study of Double-Bond Migration in Allyloxy Polyether Alkoxides Edwin C. Steiner and Roger 0 . Trucks
Contribution from The Dow Chemical Company, Edgar C. Britton Research Laboratory, Midland, Michigan 48640. Received September 1 , 1967 Abstract: The rates of rearrangement of CH2=CHCH2(0C3H&O-M+ to cis-CH3CH=CH(OC3H&0-M+ have been studied kinetically at 30" in tetrahydrofuran solution. The rates are cation dependent. The Cs+ compounds react about 60 times as fast as the K + compounds; the Na+ compounds do not rearrange under the conditions used, The rates depend on the value of n, being maximum when n = 3 in the K + series and when n = 3 or 4 in the Csf series, Chelation of cation by the polyether chain is adduced as the cause of this dependence. Addition of fluorenylpotassium or potassium dodecylbenzenesulfonate greatly reduces the rearrangement rates, sug-
gesting that mixed ion pair aggregation is occurring in these systems.
T
here has been considerable interest in the last few years in elucidating the details of the mechanism of base-catalyzed double-bond migration reactions. An excellent review of the subject is given by Cram,' who points out many of the subtleties of the reaction, including the effects of the solvent, the type of base catalyst, and the associated cation on the course of the reaction. During our study of the base-catalyzed polymerization of propylene oxide (PO),2 we have had the occasion to study the rates of rearrangement of a series of allyl ethers to the corresponding propenyl ethers (1) D. J. Cram, "Fundamentals of Carbanion Chemistry," Academic Press Inc., New York, N. Y., 1965. (2) E. C. Steiner, R. R. Pelletier, and R. 0.Trucks, J . Amer. Chem. Soc., 86, 4678 (1964).
and have found some interesting results which we would like to report here. The polymerization of propylene oxide is ordinarily carried out at elevated temperatures using basic catalysts and mono- or polyhydroxylic compounds as initiators. Price and St. PierreY3however, found that the polymerization could be accomplished at room temperature with anhydrous KOH as the catalyst and with no added hydroxylic initiators. The products of the polymerization under the latter conditions comprised mainly compounds of the type CHFXHCH~O(C~H~O).H
AO(PO)mH
(3) L. E, St. Pierre and C. C. Price, ibid., 78,2432 (1956).
Journal of the American Chemical Society / 90:3 / January 31, I968
721
and cis-CH8CH=CHO(C3HeO).H
PO(PO),H
A detailed study of the polymerization in this laboratory revealed that the ratio of propenyl to allyl ethers remained almost constant throughout the reaction, a result which was shown to be caused by a marked dependence of the rate of rearrangement of the allyl ethers on the number of PO units in the chains.* It was also found that the polymerization conditions were such that the hydroxylic end groups were almost quantitatively in the form of potassium alkoxides. These results led us to study the rearrangement of the compounds AO(PO),-M+ where n = 1-7 in order t o gain an understanding of the factors which cause the dependence of rate on chain length.
Experimental Section Solvents. Tetrahydrofuran (THF) was distilled from potassiumnaphthalene complex and stored under nitrogen at all times. The storage flask was equipped with a three-way Teflon stopcock, and the liquid was transferred cia syringe through the stopcock while nitrogen was flowing through the side arm of the stopcock to prevent entrance of oxygen or moisture. The dimethyl ether of tetraethylene glycol (Ansul Ether 181) was fractionated before use and stored under nitrogen, bp 107" (1.5 torr). Allyl Ethers of Propylene Oxide Oligomers, AO(PO),H. Allyl alcohol (212 g, 3.61 mol,
